Making sense of data

Share this on social media:

The OECD has just launched a service that allows users to navigate across all its databases. Enrico Giovannini, chief statistician (pictured left), and Toby Green, head of publishing (pictured right), explain why

What does the OECD publish?

EG: The OECD was built to foster international cooperation. We act as a facilitator, working with national experts. We have 30 member countries but also work with non-member countries.

We have data on very different subjects and also data on different territorial levels. We’ve fully convinced that we have to share this information outside the organisation.

We sell access to the big databases on annual subscription but we are very liberal with free trials and also encourage people to share subscriptions in some countries.

We are a not-for-profit publishing organisation but our member countries want us to cover the cost of publication.

Users keep whatever they had if they stop subscribing. We also supply the databases on CD-ROM annually so, that librarians can archive them.

We have just launched a service called OECD.Stat. This allows users to navigate across OECD’s databases and save searches. It’s been launched as a beta version and will be free until at least July.

We are getting a lot of feedback. In the pre-beta version we reached 16,000 hits per day. We expect that to grow nicely over the next few months with the full dataset.

How is this information used?

EG: What makes OECD a unique source is that we have not just data but also analytical products. In our publications, via DOI, we now have the possibility of linking between the data and analysis.

TG: We print DOIs under the data. Clicking on them delivers an Excel file directly to the screen with information about, for example, the relevant book. e’re currently rebuilding our platform, called SourceOECD, and the next version will enable users to click directly to the publication.

We have also developed a widget that sits in Excel files and checks whether the data has been updated since it was last used. We haven’t put that into our production plan yet but it will probably be in 2009.

EG: The next step is to enable people to make links between their analysis and our data. We’re developing that with CrossRef. We’re building a whole new way of citing databases that doesn’t exist now.

More and more people want to integrate data from different sources. All our online publications – books, journals, databases – have been on one single platform, SourceOECD, for several years.

Why integrate this information?

TG: We saw that this was the way the industry was going. It was clear to us that librarians saw us as a single information provider so we put everything on the same platform using the journal pay-once subscription model. We started making PDF files of our books in 1998 and we now have more than 4,000 books in the system. We put our books straight into the journal platform. This was a bit clumsy at first but then we started to break them down into chapters and write abstracts for each chapter. This gives the ability to hop between content.

When we started putting our books online a lot of publishers thought we were mad. However, this move saved our publishing division. We aren’t worried about the future of print. We leave it up to the librarians to decide. As long as people still want print versions we’ll print them.

What is the role of training?

EG: Our tool is very powerful but can also be very dangerous if you compare the wrong data. Almost anyone can put figures together but the amount of information produced today is huge. The challenge is to discover and select relevant information.

This is where libraries play a key role. The internet gave the impression that everyone can be experts without asking anyone but now there’s so much information that people need to ask for help again. We see this as an increasing trend. This is why we are investing in training. It’s a partnership with librarians.

What’s next?

TG: We are trying to move towards telling a story about what the data means. We give snapshots of our databases and of particular countries in our free Factbooks.

We work with swivel.com, a sort of YouTube for data, and Many Eyes, IBM’s online social tool for visualising data. We have loaded our Factbooks onto both of these. We are also going to include Podcasts on our new platform to explain the data and its implications.

Interview by Siân Harris